Microsoft Word - E0175091511

158 Abstract—This paper presents an investigation of selection strategies upon the Artificial Bee Colony (ABC) algorithm in examination timetabling problems. ABC is a global stochastic optimisation algorithm that is based on the behavior of honey bee swarms. Onlooker bees in ABC algorithm choose food source based on the proportional selection strategy. In this paper, three selection strategies are introduced (i.e. disruptive, tournament and rank selection strategies), in order to improve the diversity of the population and avoid the premature convergence in the evolutionary process. Experimental results show that the modified ABC with the three selection strategies outperforms the ABC algorithm alone. Among the selection strategies, the disruptive selection strategy shows the better performance when tested on standard benchmark examination timetabling problem.